Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: Testing the difference between coefficients with interaction variables


From   Tirthankar Chakravarty <tirthankar.chakravarty@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Testing the difference between coefficients with interaction variables
Date   Sat, 27 Aug 2011 01:00:52 -0700

But that is implicit in what you are testing (combine equation 1 and
negative of equation 2 to get first of second set). Recall that Wald
tests need the row rank of the constraint matrix to be full - the
latter two constraints are redundant and would lead to a degenerate
constraint matrix.

T

On Sat, Aug 27, 2011 at 12:43 AM, Michael Crain <michaelcrain@ymail.com> wrote:
>> /****************************/
>> sysuse auto, clear
>> regress price c.mpg##i.rep78
>> testparm c.mpg#i.rep78, equal
>> /****************************/
>
>
> The output from this code seems to be testing the joint significance of:
> group 2 - group 3 = 0
> group 2 - group 4 = 0
> group 2 - group 5 = 0
>
>
> I don't see these results testing:
> group 3 - group 4 = 0
> group 4 - group 5 = 0
>
> Am I misinterpreting?
>
>
> --- On Sat, 8/27/11, Tirthankar Chakravarty <tirthankar.chakravarty@gmail.com> wrote:
>
>> From: Tirthankar Chakravarty <tirthankar.chakravarty@gmail.com>
>> Subject: Re: st: Testing the difference between coefficients with interaction variables
>> To: statalist@hsphsun2.harvard.edu
>> Date: Saturday, August 27, 2011, 3:10 AM
>> Add the "equal" option:
>>
>> /****************************/
>> sysuse auto, clear
>> regress price c.mpg##i.rep78
>> testparm c.mpg#i.rep78, equal
>> /****************************/
>>
>> T
>>
>> On Sat, Aug 27, 2011 at 12:04 AM, Michael Crain <michaelcrain@ymail.com>
>> wrote:
>> > Yes, I see this is testing for differences between the
>> base group
>> > compared to each of the other groups:
>> >
>> > base group - group 2 = 0
>> > base group - group 3 = 0
>> > etc.
>> >
>> > I am trying to test not only base group - group 2 =
>> 0,
>> > but also: group 2 - group 3 = 0
>> >
>> > Can examining coefficients between groups other than
>> the base group be
>> > done in a straightforward way?
>> >
>> >
>> > --- On Sat, 8/27/11, Tirthankar Chakravarty <tirthankar.chakravarty@gmail.com>
>> wrote:
>> >
>> >> From: Tirthankar Chakravarty <tirthankar.chakravarty@gmail.com>
>> >> Subject: Re: st: Testing the difference between
>> coefficients with interaction variables
>> >> To: statalist@hsphsun2.harvard.edu
>> >> Date: Saturday, August 27, 2011, 2:28 AM
>> >> Here is an example:
>> >>
>> >> /****************************/
>> >> sysuse auto, clear
>> >> regress price c.mpg##i.rep78
>> >> testparm c.mpg#i.rep78
>> >> /****************************/
>> >>
>> >> This tests whether or not the coefficient on "mpg"
>> is
>> >> significantly
>> >> different across categories of "rep78". In this
>> case, you
>> >> are unable
>> >> to reject the hypothesis that they are
>> indistinguishable
>> >> from the
>> >> coefficient of "mpg" for the base group.
>> >>
>> >> T
>> >>
>> >> On Fri, Aug 26, 2011 at 11:12 PM, Michael Crain
>> <michaelcrain@ymail.com>
>> >> wrote:
>> >> > I know testing the difference between
>> coefficients of
>> >> variables is
>> >> > straightforward:
>> >> >
>> >> > reg y x1 x2 x3
>> >> > test x1 = x2 = x3
>> >> >
>> >> > Is the test as straightforward for examining
>> >> coefficients in models
>> >> > with interactions variables?
>> >> >
>> >> > reg y x1 x2 x3 x1*x2 x1*x3
>> >> >
>> >> >   where:
>> >> >
>> >> >     x2 and x3 are dummy variables that
>> identify
>> >> groups
>> >> >
>> >> >     the base group is when x2=0 and x3=0,
>> group 2
>> >> >     is when x2=1 and x3=0, group 3 is when
>> x2=0 and
>> >> x3=1
>> >> >
>> >> > I am interpreting the effect of the base
>> group as x1's
>> >> coefficient. The
>> >> > effect of group 2 is the sum of coefficients
>> of the x1
>> >> and x1*x2
>> >> > variables. Lastly, the effect of group 3 is
>> the sum of
>> >> coefficients of
>> >> > the x1 and x1*x3 variables.
>> >> >
>> >> > Now, I want to examine whether the effects of
>> the
>> >> three groups are
>> >> > different.
>> >> >
>> >> > Is the test:
>> >> >
>> >> > test x1 = x1*x2 = x1*x3
>> >> >
>> >> > *
>> >> > *   For searches and help try:
>> >> > *   http://www.stata.com/help.cgi?search
>> >> > *   http://www.stata.com/support/statalist/faq
>> >> > *   http://www.ats.ucla.edu/stat/stata/
>> >> >
>> >>
>> >>
>> >>
>> >> --
>> >> Tirthankar Chakravarty
>> >> tchakravarty@ucsd.edu
>> >> tirthankar.chakravarty@gmail.com
>> >>
>> >> *
>> >> *   For searches and help try:
>> >> *   http://www.stata.com/help.cgi?search
>> >> *   http://www.stata.com/support/statalist/faq
>> >> *   http://www.ats.ucla.edu/stat/stata/
>> >>
>> >
>> >
>> > *
>> > *   For searches and help try:
>> > *   http://www.stata.com/help.cgi?search
>> > *   http://www.stata.com/support/statalist/faq
>> > *   http://www.ats.ucla.edu/stat/stata/
>> >
>>
>>
>>
>> --
>> Tirthankar Chakravarty
>> tchakravarty@ucsd.edu
>> tirthankar.chakravarty@gmail.com
>>
>> *
>> *   For searches and help try:
>> *   http://www.stata.com/help.cgi?search
>> *   http://www.stata.com/support/statalist/faq
>> *   http://www.ats.ucla.edu/stat/stata/
>>
>
>
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/
>



-- 
Tirthankar Chakravarty
tchakravarty@ucsd.edu
tirthankar.chakravarty@gmail.com

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index